I'm curious what sorts of profit factors people look to get on the initial idea for a strategy they think has potential. My gut tells me that it's rare to find a strategy which, prior to optimization, has a profit factor above 2, but my gut could be completely off on this. In my case, although I only use one strategy across multiple markets, I have found that the "base case" for the strategy results in a profit factor of between 1 and 1.4. In one market, it's slightly below 1, but not by much. Anyone ever get an idea, test it and find that even before you do anything to optimize, you've got a profit factor of higher than 2? On the other side, what would be the lowest initial profit factor you would think could be worked into something useful with optimizations? 0.9? 0.75? Nothing below 1, ever?
The initial profit factor is meaningless, as it results from random parameter values. It can even be higher than after optimization. I think only after optimization, when you see how the profit factor behaves along parameter ranges, you can see whether the strategy is worth to be pursued further.
That's interesting that you assume that the initial parameter values are random. Do you think that is the case for all strategies? I ask because I don't consider my initial parameter values random at all, or, at least not my primary parameter. As I gained more experience with the strategy, I found that other parameters, which did seem random initially, also appear to have an impact on outcomes.
Before you optimize a strategy, how do you know how your parameters affect the PF? Of course you select some values that you think make sense, but they are normally very different from the values returned by the optimizer. I had sometimes strategies where my initial settings, by chance, gave better PF results than the optimized values - they just happened to be at a performance peak. The optimizer I'm using avoids narrow peaks and looks for flat hills.